Recent transparency policies across several U.S. states, like salary history bans that prevent employers from requiring candidates to report their compensation or even their salary expectations, have heightened the focus on fair salary practices in the workplace. As a result, there’s an increased demand for strategies to determine appropriate compensation. A key solution lies in salary benchmarking: using aggregated market data to establish competitive pay rates.
Setting salaries is a balancing act: you do not want to offer salaries that are too high relative to the market, because that’d be wasteful; but offering salaries that are below market may make it difficult to attract, retain and motivate your employees. How do employers find their sweet spot?
In a recent study, we developed a theoretical model to shed light on the demand for salary benchmarking. In our model, firms know exactly how much value the job candidate would bring to the table. The firm would like to offer the candidate below that value to keep some of the value as profit. However, how much lower should the offer be? The lower the offer, the better the savings. On the other hand, if the offer is too low, the firm risks losing the job candidate to another firm — they may turn down the offer, or accept the offer and leave shortly thereafter.
So, whether the company wants to make a more or less aggressive offer depends critically on what the firm believes is the market value of the job candidate: i.e., how much other firms are willing to pay for the job candidate. That’s where salary benchmarking comes in: firms can use data on the offers accepted by similar employees (e.g., with the same position title and industry) to make sure that their offer is not too low, or too high, relative to the market benchmark.
Our research also provides empirical evidence that access to salary benchmarks has a significant effect on the salaries that firms set. We conducted it in collaboration with the largest U.S. payroll processing company, ADP, serving 650,000 firms and processing paychecks for 20 million Americans. In addition to the payroll services, the company aggregates the salary data from their payroll records in the form of salary benchmarks. Clients can then search for any job title they want in a user-friendly way. It is one of the most advanced benchmark tools and is being used by many prominent firms.
Our data covers the roll-out of the tool when it was first introduced to the market, late 2015, through the start of the pandemic in 2020. Our sample includes 584 “treatment” firms that gained access to the tool and 1,431 “control” firms that did not gain access to the tool but were selected because they shared similar characteristics to the treatment firms. We studied the evolution of salaries in the treatment firms shortly before and after they gained access to the benchmarking tool, and we compared that to the corresponding evolution of salaries in the control firms.
Here’s what we found: let’s say that a firm sets the “right” salary when they choose a salary that is in a narrow (5%) band of the median salary for the same position, industry, and state. Prior to gaining access to the benchmarking tool, there was an 11.6% probability that a firm sets the “right” salary. After gaining access to the benchmarking tool, the probability of setting the “right” salary increased two-fold, to 22.1%. This is compelling evidence that these salary benchmarks are truly influencing pay-setting, and causing position-level salaries to converge across companies.
We also found that the effects of salary benchmarking are quite symmetric. After looking up the benchmark, firms are less likely to pay significantly above-market, presumably because it would be wasteful; but firms are also less likely to pay significantly below-market, presumably because it would be costly in terms of attracting, retaining and motivating employees.
For standardized positions, such as bank teller and receptionist, we found some evidence of a modest increase in the average salary. If employers are choosing to raise salaries for this group, it must be because they expect to get something in return, right? For instance, bumping salaries up to market may allow employers to retain their employees and save those pesky turnover costs. Our research shows that, indeed, benchmarking improves retention. Among employees in these roles, the gain of about 6% in average salary was followed by an increase of about 16% in the retention rate over the following 12 months.
How can I use salary benchmarking?
The payroll data show that high quality salary benchmarks sway employers to set pay closer to the marketplace median pay, and that benchmark users benefit from a boost in employee retention. But how exactly does one go about sourcing and using a salary benchmark?
Salary benchmarks are wide ranging; some are created by the government and others exclusive to the clients of consulting firms. Still others are crowdsourced data from anonymous employees. For example, the U.S. government reports average earnings at the six-digit occupation code, within industry and state, each year using surveys of businesses and households. These data are highly representative, but are released infrequently and at a higher level of aggregation than the sought after position level information. Meanwhile, several large consulting firms offer exclusive access to the aggregated earnings of their clients, often using proprietary standardized corporate titles. Most recently, firms that collect salary information as part of their core businesses, like ADP, have added user interfaces that make these data accessible as benchmarks. And platforms like Glassdoor have encouraged millions of users to volunteer their salary information anonymously in exchange for seeing the average inputs of others at the position level.
To better understand how firms use salary benchmarking, we also conducted a survey with the members of the Society for Human Resource Management Research Institute (SHRM). We collected responses from a large sample of 2,085 HR managers in charge of pay-setting. This sample spans organizations of every size, state and industry, including both public and private sectors. To complement this data, we conducted an additional survey with a sample of clients from ADP, in which we obtained responses from 720 HR managers in charge of pay-setting.
Employers use salary benchmarking for a wide range of purposes. For instance, 54% of respondents use benchmarking when setting the pay of new hires. Other popular uses include salary negotiations (53%), setting ranges for specific job titles (90%), and setting the salary for a job advertisement (41%). Employers also vary in terms of how frequently they consult salary benchmarks. For instance, while some employers do benchmarking for every single new hire (37% of respondents), the rest prefer to do it for some specific new hires or for some specific positions.
Our survey data indicates that employers differ in which data sources they use. The most popular option is to use industry surveys, chosen by 68% of respondents, but hiring a consulting firm to conduct these surveys can be quite expensive, and there is no guarantee that the survey data will live up to the expectations. If you do not have a big budget, you may also want to try the second most popular option: free online data sources (58% of respondents) such as Glassdoor.com. More recently, payroll service companies have gained popularity (used by 23% of respondents and the fastest growing source), which may give the best bang for your buck in terms of data quality. Most employers, however, use a combination of data sources. While the majority look at free online sources, on a scale of 1-5, HR professionals report that other sources are twice as trustworthy as free online sources.
Labor market trends can sometimes change rapidly, especially in the post-pandemic world. You do not need to check the benchmarks on a daily basis, but try to at least do it from time to time — remember that information, too, has an expiration date. Likewise, you may not have time to sit down to study market trends for every position in which you hire. However, make sure to at least do due diligence for the positions that are most critical for your business.
When using a benchmarking tool, you may feel information overload. Some tools provide not only the median pay, but also the mean, 10th percentile, 25th percentile, and so on and so forth. And some tools allow you to filter by industry, state, and other criteria. There are trade-offs. For instance, filters can allow you to focus on peer employers, but at the cost of smaller sample sizes and thus statistically imprecise benchmarks. While the right approach depends on the context at hand, according to our survey, there are some clear favorites: most firms (89%) prefer to look at the median market pay or the average market pay, and the most popular filters are by industry (87%) and state (84%).
We also gave the survey respondents an opportunity to share their own tips and advice for salary benchmarking. Some recommended not relying on job title information alone, but rather to pay attention to the job description to ensure it’s a good match. Another manager suggested that, to the extent possible, you should review the benchmarks regularly to stay in sync with the latest trends in the job market. In the same spirit, one manager recommends being proactive — do not wait until employees are unhappy about low pay and reach out to you, or you risk higher turnover.
In sum, the pay transparency revolution extends to firms knowing much more about what their competitors are offering. New high-quality sources have changed pay-setting practices. According to our latest study, keeping up with pay in the marketplace, and knowing how to use those sources, can affect your bottom-line retention and payroll costs. And from a societal perspective, this is closing the gap between pay across firms.